Here’s an interesting abstract from the Winter 2005 edition of the Harvard International Journal of Press/Politics:

This article reports on a cross-cultural analysis of television coverage of the 2003 Iraq War that seeks to assess and understand the dimensions of objectivity in the news during wartime. A total of 1,820 stories on five American networks (ABC, CBS, NBC, CNN, Fox News Channel [FNC]) and on the Arab satellite channel Al Jazeera were included in the study. The study assessed bias on two levels: tone of individual stories and the macro-level portrait of the war offered by each network. Results showed that at the story level, the overwhelming number of stories broadcast by Al Jazeera and the American networks other than FNC were balanced.

OK, so Al Jazeera and the network news are balanced, but Fox is biased. Check. Even if this study is true, do the networks really want to be in the same boat as Al Jazeera?

I did a quick read of the article and found some elementary questions about its methodology. The authors allowed individual coders to judge the “tone” of each story on a 5-point scale. One of the guidelines to bias was whether the reporters used the term “our troops” when talking about coalition forces. That criteria alone would produce the “Fox is biased” result.

(Very inside baseball critique: Also, the authors wisely used more than one coder, so they could compare the results of each coder to check for reliability. However, the authors only had one Arabic coder who could understand the Al Jazeera broadcast, so there was no intracoder reliabiltiy for one of the most important elements of the study.)

In short, the results are debatable. I can think of far better ways to conduct a content analysis than giving coders a 5-point scale and letting them guess. But, the authors may have had their result in mind before they started the process. Perhaps they’ll say the same about the results of my content analyses. I wonder if I’ll find mine as easy to publish?

Perhaps we’re going to start seeing more and more outright partisanship in peer-reviewed journals. Don’t forget this widely cited study (in the Lancet) which found 100,000 Iraqis died in the war. It has been widely debunked.